Upload
others
View
10
Download
0
Embed Size (px)
Citation preview
Kinematic Analysis of Sit-To-Stand Assistive Device for the Elderly and Disabled
Inho Kim, Woonghee Cho,Gyunghwan Yuk Hyunseok Yang
Department of Mechanical Engineering Yonsei University
Seoul, Korea
Byeong-Rim Jo, Byung-Hoon Min Mechatronics & Storage Laboratory
LG Electronics, Inc Seoul, Korea
Abstract—This paper introduces a robotic walking and sit-to-stand support system, Smart Mobile Walker(SMW), and presents kinematic analysis of the system. For supporting person’s sit-to-stand movement, the motion of SMW is modeled by the analysis of kinematics. The results of analysis are used for simulations and its feasibility is investigated by simulations. The feasibility shows our analysis of kinematics can be applied as a base of making optimal motion trajectory for helping sit-to-stand.
Keywords – Rehabilitation robotics; Assistive technology; Sit to Stand; Analysis of Kinematics; Motion Trajectory
I. INTRODUCTION Recently, the number of elder people is increasing and thus
interest of welfare for the elderly and disabled are rapidly rising as well. This rapid increase of attention implies that elders require more technical aid to support their daily living. Especially, numerous problems can be seen in the essential activities such as sitting down and standing up. Therefore, many recent studies focus on solving those problems.
Researches about sit-to-stand can be categorized into two parts; perspective of analyzing sit-to-stand motion and designing device.
In the analyzing sit-to-stand motion, a trajectory of physical points such as head, arms, trunk, hip, knee and ankles has been studied. One approaches to design and validate of a three-segment human body model [4]. And some researches tried to find optimal trajectories using cost function [5].
In the designing device, some trials of assistive devices for sit-to-stand movement have been reported. The powered knee-orthosis have been proposed to help user’s sit-to-stand movement [2]. And passive gravity balanced assistive device introduced [1]. This device use gravity for reducing joint torque when sit-to-stand and a hybrid method is used to manipulate center-of-mass of the system. There is a device supporting user’s hip using moving bed and handrail [3]. And walker type which is similar with normal walker is introduced [8]. Lately, a robot which can help multimodal walking and sit-to-stand motion has been proposed. An assistive robotic device,
KineAssist, [6] has been applied for therapy of rehabilitation as like gait and balance training.
In [1], the device is proposed to support user’s hip for helping his sit-to-stand movement. It decreases required torque at the knee and efficiently supports the upper body. In [6], the device was designed to support user’s trunk, because this paper insists that supporting trunk is the important factor in sit-to-stand movement. Similarity of above two papers is the support of upper body, which is a major factor in sit-to-stand. Therefore, SMW is designed to support arms in order to control upper body.
Section 2 briefly introduces our device, SMW. Section 3 proposes analysis of forward and inverse kinematics. Section 4 describes results from verification. Section 5 presents conclusion including remarks and future works.
II. SMART MOBILE WALKER(SMW) SMW has been developed as a device for helping the
elderly and disabled. Especially, SMW’s development focuses on giving a hand to their daily living like walking and sit-to-stand movement. Thus, we had designed the SMW in order to support user’s walking and sit-to-stand movement.
Figure 1. CAD of Prototype of SMW
The SMW is composed of six linear actuators, two hub motors, two passive wheels, two servo motors and a top plate. Each servo motor controls steering angle of the hub motors for
2011 IEEE International Conference on Rehabilitation Robotics Rehab Week Zurich, ETH Zurich Science City, Switzerland, June 29 - July 1, 2011
978-1-4244-9861-1/11/$26.00 ©2011 IEEE 692
walking support. Four linear actuators, linking legs and the body, control orientation and position of the top plate. Remaining two linear actuators, linking the top plate and the body, create up and down motion to the top plate.
Figure 2. The Sit-to-Stand Motion with Top Plate
Sit-to-stand assistance using SMW can be performed by controlling the linear actuators. The angle and height of the top plate change following the linear actuators. Therefore kinematic analysis is essential to get the relationship between top plate condition and linear actuators’ length.
III. KINEMATIC ANAYSIS Sit-to-stand assistance rarely needs different motion on left
and right side of SMW. Thus kinematics model can be simplified to half body model. There are three linear actuators and a servo motor and a hub motor on one side. However degree of freedom required to the top plate is 3. Therefore we need to fix two actuators to have unique inverse kinematics solution. Steering servo and linear actuator #3 are to be fixed for the reason.
A. Forward Kinematics Forward kinematics of SMW is defined as figuring out
position and angle on the top plate from given linear actuator lengths. The hub wheel motion is directly related to the x coordinate of the top plate. Thus it is not under consideration on kinematic analysis. Then the output parameter is simplified as the height and angel of the top plate.
Figure 3. Simplified SMW for forward kinematics
Fig.3 shows the simplified model for forward kinematics. There are two variables, length of the front actuator and rear actuator. Angles between the body and legs are decided by linear actuators lengths as equation (1).
2 2 2
2 2 2
acos2
(1)acos
2
s a rr s
s a
s a ff s
s a
L L LL L
L L LL L
θ θ
θ θ
⎫⎛ ⎞+ −= − ⎪⎜ ⎟
⎝ ⎠ ⎪⎬
⎛ ⎞+ − ⎪= −⎜ ⎟ ⎪⎜ ⎟⎝ ⎠ ⎭
Figure 4. Part of SMW for Analyzing Forward Kinematics
As the length of the legs and body size is known, we can specify a quadrilateral formed by drawing a line connecting the wheel axis of hub wheel and passive wheel as Fig.4. The length of the line can be calculated as equation (2), by assuming the rear leg is overlapped on X axis of local coordinate system and the origin is on passive wheel axis.
( ) ( ) ( )( ) ( ) ( )
( ) ( )2 2
2 (2)
2
θ
f r l r r r f
f r l r r r f
r fk
f r
rR w f r f rw
x x L L cos Dsin Lcos
y y L sin Dcos Lsin
y yatan
x x
R rasin L x x y yL
θ θ θ θ
θ θ θ θ
θ
⎫= + + + + + ⎪⎬
= − + − + ⎪⎭⎛ ⎞−
= ⎜ ⎟⎜ ⎟−⎝ ⎠⎛ ⎞− ⎛ ⎞= = − + −⎜ ⎟ ⎜ ⎟
⎝ ⎠⎝ ⎠
From the length of the line and radius of wheels, angle
rRθ can be decided as above. Then the angle of the top plate is decided as equation (3).
( ) ( )2 2
( ) (3)
rR w f r f rw
t k rR r rt k rR
R rasin L x x y yL
θ
θ θ θ θ θ θ θ
⎛ ⎞− ⎛ ⎞= = − + −⎜ ⎟ ⎜ ⎟⎝ ⎠⎝ ⎠
= + − = +
693
( ) ( ) ( ) ( )( ) ( ) ( ) ( )
(4)t r rt l t t t
t r rt l t t t
x x Lcos L cos H D sin
y y Lsin L sin H D cos
θ θ θθ θ θ
= + + − + ⎫⎪⎬
= + + + + ⎪⎭
The height of the top plate can be represented as equation (4) using trigonometric functions. It assumes that the y axis coordinate at ground is zero. X axis coordinate has zero at the center of passive wheel.
B. Inverse Kinematics Inverse kinematics is opposite to forward kinematics. The
information of top plate becomes input data and output are length of actuators. We define new points and model constant lengths for analyzing inverse kinematics. The simplified model is shown in Fig. 5.
Figure 5. Simplified Model for Inverse Kinematics
There are three variables; x-coordinate, y-coordinate of top plate and the angle of top plate. These factors are used as input data. If the information of top plate were determined, lengths of actuators would be calculated. In this time, we use sets of six points for getting length of actuators. First, we define coordinates of four points.
sin cos
cos sin
sin coscos sin
( )sin cos
( ) cos sin
fh t f t h tfh
fh t f t h t
rh t r t h trh
rh t r t h t
f t f s t l tf
f t f s t l t
x x H LP
y y H L
x x H LP
y y H L
x x H H LP
y y H H L
θ θθ θ
θ θθ θ
θ θθ θ
= + + ⎫⎪⎬= − + ⎪⎭
= + − ⎫⎬= − − ⎭
= + + + ⎫⎪⎬= − + + ⎪⎭
( )sin cos( ) cos sin
r t r s t l tr
r t r s t l t
x x H H LP
y y H H Lθ θθ θ
= + + − ⎫⎬= − + − ⎭
From above result, we can derive two points where equip the actuator on the each leg. For this, we define another variable as indicated in Fig. 6.
Figure 6. Simplified Model for Inverse Kinematics
From the point information, lengths of each actuator can be found as equation (5).
( ) ( )( ) ( )
2 2
2 2
sin cos sin cos
( ) y
( ) y
(5)
f rf r
f a f afl f fl f
r a r arl r rl r
f fh fl fh fl
r rh rl rh rl
y R y rS L a S L a
L L
S L y R Lx x y
L LS L y r L
x x yL L
L x x y y
L x x y y
⎛ ⎞⎛ ⎞− ⎛ ⎞⎛ − ⎞⎜ ⎟⎜ ⎟= = ⎜ ⎟⎜ ⎟⎜ ⎟⎜ ⎟⎜ ⎟ ⎝ ⎠⎝ ⎠⎝ ⎠⎝ ⎠−
= + = +
−= − = −
⎫= − + − ⎪⎬⎪= − + − ⎭
Ultimately, length of actuators can be calculated from the information of top plate. There is no doubt that if front wheel runs forward or backward, it would change x-position of the top plate. However since we focused on sit-to-stand motion this time, we can neglect the movement of front wheel.
IV. VERIFICATION In this section we will discuss the feasibility of SMW. The
feasibility is investigated by simulation, which is performed by taking 2 steps. First step is finding length of actuators from arbitrary information of top plate by using inverse kinematics. Second step is getting information of top plate using length of past actuators by forward kinematics. Fig.7 represents the schematic diagram of simulation procedure.
694
Figure 7. Schematic Diagram of Simulation Procedure
In first step, certain reference trajectory of the top plate was taken as Fig. 8.
Figure 8. Reference Trajectory of Top Plate
Fig. 8(a) is position of top plate and Fig. 8(b) is orientation of top plate. We apply inverse kinematics to this reference trajectory for getting length of actuators. After getting the length of actuators, we used it as input data for forward kinematics. Through forward kinematics, we can obtain information of the top plate. Fig.9 shows error between reference trajectory and information of top plate from forward kinematics.
0 100 200 300 400 500 600 700 800-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
1
time(ms)
Err
or
Figure 9. Error between Reference and Result of Forward Kinematics
There are no errors between reference and result of forward kinematics. This means that if there were a specific motion trajectory which needs to be obtained, it is possible to realize it by using analysis of kinematics.
V. CONCLUSION & FUTURE WORK This paper analyzed kinematics of the robotic walking and
sit-to-stand support system, SMW. Its feasibility is investigated by simulation. Moreover, a possibility that kinematics can be applied in order to develop optimal motion trajectory was verified. Therefore, next research can be made focusing more on generating optimal motion trajectory, expected to experiment to gather information of trajectories and evaluate them. This research also expects to apply physical theories and develop the algorithm for making optimal motion trajectory.
ACKNOWLEDGMENT This work was supported by the R&D Program of
MKE/KEIT[10035201, ADL Support System for The Elderly and Disabled].
REFERENCES [1] Abbas Fattah, Sunil K. Agrawal, Glenn Catlin, John Hamnett, “Design
of a Passive Gravity-Balanced Assistive Device for Sit-to-Stand Tasks” Journal of Mechanical Design, vol. 128, pp. 1122–1129, September 2006.
[2] Wesley R. Eby, Eric Kubica, “Modeling and Control Considerations for Powered Lower-Limb Orthoses : A Design Study for Assisted STS”, Journal of Medical Devices, vol. 1, pp. 126-139, June, 2007.
[3] Yasuhisa Hirata, Jun’ichi Higuchi, Takuro Hatsukari, Kazuhiro Kosuge, “Sit-to-Stand Assist System by Using Handrail and Electric Bed Moving Up and Down.”, Proceedings of the 2nd Biennial IEEE/RAS-EMBS International conference on Biomedical Robotics and Biomechatronics, pp.187-192, 2008
[4] Josip Music, Roman Kamnik, Marko Munih, “Model based inertial sensing of human body motion kinematics in sit-to-stand movement”, Simulation Modelling Practice and Theory, vol. 16, 933-944, 2008
[5] Jernej Kuzelicki, Milos Zefran, Helena Burger, Tadej Bajd, “Synthesis of standing-up trajectories using dynmic optimization”, Gait and Posture, vol. 21, pp.1-11, 2005
[6] Michael Peshkin, David a. Brown, Julio J.Santos-Munne, Alex Makhlin, Ela Lewis, J. Edward Colgate, James Patton, Doung Schwandt, “ KineAssist: A robotic overground gait and balnce training device”, Preceedings of the 2005 IEEE 9th Interantional conference on Rehabilitation Robotics, pp.241-246, June, 2005
[7] Makoto Suzuki, Ken Masamune, Lin-Jong Ki, Takeyoshi Dohi, Hideo Yano, “Development of a robot arm controlled by force sensors as a walking aid for elderly”, Robotica, vol. 16, pp. 537-542, 1998
[8] Oscar Chuy Jr., Yahuhisa Hirata, Zhidong Wang, Kazuhiro Kosuge, “Approach in Assisting a Sit-to-Stand Movement Using Robotic Walking Support System”, Proceedings of the 2006 IEEE/RSJ International Conference on Intelligent Robots and Systems, pp. 4343-4348, October, 2006
TABLE I. PARAMETERS IN ANALYSIS OF KINEMATICS
Symbol Forward & Inverse Kinematics Definition
tx The x-coordinate of top plate
ty The y-coordinate of top plate
tθ The angle between top plate and horizental axis
rL The length of rear linear actuator
695
fL The length of front linear actuator
r The radius of rear wheel R The radius of front wheel
sL The distance between linking point of body and actuator and leg’s joint
aL The distance between leg’s joint and linking point of actuator and leg
lL The distance between leg’s joint and the middle of body
L The distance of leg
Symbol Forward Kinematics Symbol Inverse Kinematics
Definition Definition rx The x-coordinate of
rear wheel rx The x-coordinate of
rear joint ry The y-coordinate of
rear wheel ry The y-coordinate of
rear joint fx The x-coordinate of
front wheel fx The x-coordinate of
front joint fy The y-coordinate of
front wheel fy The y-coordinate of
front joint tH The distance between
top and front leg’s joint
rH The distance between top and linking point of actuator and body on rear side
D The difference between rear wheel and front wheel
fH The distance between top and linking point of actuator and body on front side
wL The distance between center of rear wheel and front wheel
sH The distance between leg joint and linking point of actuator and body
sθ The angle between hL The distance between
middle of body and
sL and inclined axis linking point of actuator and body
rθ The angle between rear leg and inclined axis
rhx The x-coordinate of linking point of actuator and body on rear side
fθ The angle between front leg and inclined axis
rhy The y-coordinate of linking point of actuator and body on rear side
kθ The angle between rear leg and line from center of rear wheel to center of front wheel
fhx The x-coordinate of linking point of actuator and body on front side
rtθ The angle between rear leg and horizental axis
fhy The y-coordinate of linking point of actuator and body on front side
rRθ The angle between line horizental axis and line from center of rear wheel to center of front wheel
rlx The x-coordinate of linking point of actuator and rear leg
Symbol Inverse Kinematics rly The y-coordinate of
linking point of actuator and rear leg
Definition
rS The distance between rear leg joint and center of rear wheel on horizental axis
flx The x-coordinate of linking point of actuator and front leg
fS The distance between front leg joint and center of front wheel on horizental axis
fly The x-coordinate of linking point of actuator and front leg
696